Simultaneous Quantitation of Morphine and Paraben Preservatives in Morphine Injectables

KEVIN L. AUSTIN and LAURENCE E. MATHERX Received June 2, 1977, from the Department ( I / Anaesthesia and Iriti~nsiiw Cart,, Flindcrs A4edicaI Centre, Flinders Uniuersity of South Australia, Rcdford Park, S.A. 5042, Australia. Accepted for puhlication February 28, 1978.

Abs t rac t 0 A high-performance liquid chromatographic method for the simultaneous determination of morphine sulfate, methylparahen, and propylparaben in morphine sulfate injection was developed. A re- versed-phase system, based on an oct.adecylsilane stationary phase, was used with a binary solvent mobile phase consisting of methanol-phos- phate buffer (pH 4.0) containing methanol (5'%) delivered a t a constant rate (0.6:0.4 ml/min) using a two-pump system. T h e detector response at 254 nm was linear with the amount injected over a wide range, allowing rapid and reproducible quantitation of each component.

Keyphrases 0 Morphine sulfate-high-performance liquid chroma- tographic analysis simultaneously with parabens in dosage form Methylparahen-high-performance liquid chromatographic analysis simultaneously with morphine sulfate and propylparaben in dosage form 0 Propylparahen-high-performance liquid chromatographic analysis simultaneously with morphine sulfate and methylparaben in dosage form 0 High-performance liquid chromatography-simultaneous analyses, morphine sulfate and parabens in dosage form Narcotic analgesics- morphine sulfate, high-performance liquid Chromatographic analysis simultaneously with parabens in dosage form

Although neither the USP (1) nor the BP (2) requires the determination of preservative agents in official prep- arations of morphine sulfate injection, a simple method for the simultaneous quantitation of morphine and pre- servative agents occasionally is required. Official methods for the determination of morphine in pharmaceutical preparations are tedious and time consuming (3). Although column chromatographic-UV spectrophotometric meth- ods were reported for the determination of paraben pre- servatives (4) and morphine in the presence of paraben preservatives (5), they also are time consuming. During a project to assess the effects of sterilization procedures on drugs used in anesthetic practice, an extremely simple and fast method was developed to quantitate simultaneously morphine, methylparaben, and propylparaben in mor- phine injectables using high-performance liquid chroma- tography (HPLC).

E X P E R I M E N T A L

Reagents a n d Chemicals-Analytical reagent grade chemicals were used. Methanol and aqueous solutions were filtered through a 0.45-pm filter' and degassed using a n ultrasonic bath immediately before use.

Morphine sulfatez, methylparaben?, and propylparabenz, all BP grade, were prepared for calibration by dissolving in distilled water, both singly and in admixture, in concentrations similar to the final concentration in the injectahle-uiz., morphine sulfate, 10.0 mg/ml; methylparaben sodium, 0.63 mg/ml; and propylparaben sodium, 0.33 mg/ml.

Apparatus-A high-performance liquid chromatograph:? was equipped with an octadecylsilane column4 for reversed-phase chroma- tography and a recording digital integrator5. The chromatograph detector

1 Celrnaii Metricel GA-6.

,' Waters ALC/GPC 200 series with CJfiK universal injector, model 440 multiple wavelength absorbance detector, model Mfi000A pumps, and model 6fi0 solvent programmer.

4 UBondapack C,". 10-pm particle size, 4 rnrn i.d. X 90 cm. 5 Hewlett-Packard model 3:WA.

Courtesv of David Bull Laboratories, Melbourne, Australia.

was set a t 254 nm. The absorbance scale was set a t 1.0 unit full scale, and the recorder was se t a t 1 crn/min.

Mobile Phase-The mobile phase was delivered from two pumps. Pump A delivered pure methanol. Pump B delivered an aqueous solution of 0.1% monobasic sodium phosphate dihydrate containing 5% methanol a t p H 4.0 (the p H was adjusted by dropwise addition of either 1% phos- phoric acid or 1% dibasic sodium phosphate). A total flow rate of 1.0 ml/min consisted of 0.6 ml/min from Pump A and 0.4 ml/min from P u m p B.

Determina t ion of Cal ibra t ion Curve-Triplicate volumes of each standard solution (1.0-10.0 pl), first individually and then in admixture, were injected into the chromatograph using a 10-pl microsyringe. Re- gression of the detector response for each component on volume injected was assessed in two ways: by manually measured peak heights and by electronically computed peak areas.

Determina t ion of Components i n M o r p h i n e S u l f a t e Injection- Exactly 3.0-p1 aliquots of morphine sulfate injection (nominally con- taining 10.0 mg of morphine sulfate/ml, 0.63 mg of methylparahen so-

l .o

- E, d In N

w $ 0.5

K

m a

- !s E: 1 3

4

I c T I 1 1 I 1 I ? l ? l

8 10 1 2 4 6 MINUTES

F i g u r e 1-High-pressure liquid chromatogram obtained with the de- scribed method. Key: peak 1, point of injection; peak 2, trace compo- nents found i n morphine sulfate injections; peak 3, morphine; peak 4, methylparahen; and peak -5, propylparaben.

1510 I Journal of Pharmaceutical Sciences Vol. 67, No. 1 1 , November 1978

0022-3549f 781 1100-1510$01.00/ 0 @ 1978, American Pharmaceutical Association

Table I-Regression Analysis of Detector Response (254 nm) Measured as Peak Height and Peak Area on Volume Injected

Methyl- Propyl- Morohine oaraben Darahen

Peak height Intercept, mm 7.4 -2.5 -1.2 Slope, mm/pl 11.71 34.21 10.25 r2 0.99406 0.99915 0.99731 Range, pl 1.0-10.0 1.0-7.0O 1.0-10.0 n 30 21 30

Intercept, (mv)(sec) - 1933 3840 -1386 Slope, (mv)(sec)/pl 20,769 31,285 15,214 r2 0.99984 0.99581 0.99826

Peak area

Range, pl n

1.0-10.0 1.0-6.0b 1.0-10.0 3n 1 R ?n

The response beyond 7.0 pl was off scale. * The response was nonlinear beyond

dium/ml, and 0.33 mg of propylparaben sodium/ml) were injected directly into the chromatograph. Peak area calibration curves were used to read off the equivalent volume injected (and, therefore, the drug amount in- jected). The results were calculated using:

6.0 PI.

(Eq. 1) percent of strength claim = - X 100%

where Vo is the volume observed as equivalent from standard curve and V , is the volume actually injected.

Effects of Sterilization on Active Ingredients-Ampuls from a test batch of morphine sulfate injection were obtained both before and after routine factory sterilization. Representative samples from each lot were divided into two groups and then subjected to three cycles of sterilization by autoclaving a t 120-121” for 30 min. One group from each lot was cooled rapidly over 5-10 min while the other group was cooled slowly over 45-60 min.

Analysis of variance was used to detect any differences between the procedures.

VO VI

RESULTS AND DISCUSSION

Mobile Phase-HPLC provides an excellent method of separation and quantitation of these compounds, which differ markedly in volatility and polarity. The analysis conditions finally chosen for routine use ful- filled the requirements of simplicity, speed, and separation. By varying the binary mixture, other drugs including lidocaine, mepivacaine, bu- pivacaine, and meperidine may readily he analyzed using the same combination of column and solvent systems. The separation of a wide range of chemical types using an octadecylsilane stationary phase with a methanol-water mobile phase has been reported (6).

Optimum operating conditions depend on the mobile phase pH. Op- eration a t pH 4.0 produced symmetrical peaks for all components (Fig. 1). Preliminary studies using an unbuffered aqueous phase and aqueous phosphate buffer a t pH 6.0 or 8.0 or an ammonium carbonate solution produced insufficiently symmetrical peaks for reproducible quantitation. Hays et al. (7) reported unsuccessful quantitation chromatography of morphine under a variety of mobile and stationary phase conditions because of peak asymmetry.

Liquid chromatography of weak bases such as morphine (pKa = 8.0) presents special problems because of the pH-dependent dissociation of the conjugate acid. This dissociation may be overcome using ion sup- pression techniques by operating a t a pH where the free base is prepon-

Table 11-Reproducibility of 10 Replicate Simultaneous Determinations of Morphine, Methylparaben. and Proovloaraben

Variable Mean CV, %

Morphine Retention time, min 3.57 1.2 Peak heinht, mm 53.70 1.4 Peak area, (mv)(sec) 71,924 0.7

Retention time, min 5.27 0.4 Peak height, mm 120.70 0.7 Peak area, (mv)(sec) 118,352 0.6

Retention time, min 8.25 1.3 Peak height, mm 33.80 1.2 Peak area, (mv)(sec) 51,320 0.6

Methylparaben

Propylparaben

Range

3.53-3.64 53.0-54.7

71,287-73,156

5.22-5.29 119.5-121.7

117,687-119,521

8.11-8.37 33.2-34.5

50,894-51,799

Table 111-Concentrations of Morphine and Paraben Preservatives in Ampuls a f t e r Three Cycles of Pharmacy Autoclave Sterilization

Not Factory Sterilized, Factory Sterilized, Mean f SD, mg/ml Mean f SLJ, mg/ml -

Control Morphine sulfate 11.31 f 0.23 11.30 f 0.45 Methylparaben sodium 0.76 f 0.03 0.76 f 0.04 Propylparaben sodium 0.38 f 0.01 0.38 f 0.00

Morphine sulfate 11.18 f 0.50 10.92 f 0.;j2

Propylparahen sodium 0.37 It 0.01 0.39 f 0.00

Rapid cool

0.76 f 0.04 Methylparahen sodium 0.77 f 0.05

Slow cool Morphine sulfate 10.84 f 0.35 11.35 f 0.34

0.77 f 0.06 Methylparaben sodium 0.76 f 0.05 Propylparahen sodium 0.37 f 0.01 0.38 f 0.01

derant. For morphine, operation in excess in pH 10 would he required, but it is not feasible because stationary phase decomposition occurs at pH 8 or greater. Recently, ion-pair techniques were developed to enable optimum separation of symmetrical peak shapes. The separation pres- ently described separates the hydrophilic conjugate acid of morphine from the more hydrophobic neutral parahen esters. Under these condi- tions, morphine is retained on the column ( k ’ = 0.4) long enough for highly reproducible separation and quantitation.

While determining optimum separation conditions, a problem of spurious detector response was encountered consistently at fractional flows of methanol between 20 and 80%. Both random fast and systematic slow noise signals were observed. This effect was attributed to the lib- eration of gas on mixing of the solvents. The liberated gas affected the constant flow rate solvent delivery system and occurred even with freshly degassed methanol and aqueous solutions. The problem was circum- vented by the addition of 5% methanol to the aqueous phase prior to degassing.

Calibration Curve and Reproducibility-There is a longstanding dispute over the choice between peak height and peak area measurements in chromatographic determinations. For two of the three components, marginally higher correlation coefficients were oh’tained from analysis of the regression of peak area on volume injected when compared to the corresponding analysis using peak heights (Table I). A definite discon- tinuity appeared in the peak height calibration curve for morphine, which was not reflected in the peak area curve. Although injection of volumes greater than 7.0 pl of methylparahen resulted i n peaks being off scale, they were still integrated. The peak area-volume injected relationsliip was linear to 6-7 pl injected; beyond this limit, distinct nonlinearity was observed. Both the peak height and peak area uersus volume injected standard curves were linear for propylparahen over the 0-10-pl range tested.

Details of a reproducibility study for 10 replicate determinations are given in Table 11. The method appears quite robust in that retention times of the compounds were very consistent with coefficients of variation in the order of 1% for all three components.

Excellent quantitative reproducibility was obtained using both peak height and peak area measurements. Coefficients of variation for peak height measurements were approximately double those for computed peak areas, confirming the preference for the latter method of calcula- tion.

Effect of Sterilization Procedure-There were no significant dif- ferences between the two rates of cooling with respect to active ingredient concentration. No degradation of active ingredients was observed for any component measured (Table 111).

REFERENCES

(1) The “United States Pharmacopeia,” 19th rev., Mack Publishing

(2) “British Pharmacopoeia 1973,’’ Her Majesty’s Stationery Office,

(3) V. D. Gupta,J. Pharm. Sci., 65,1967 (1976). (4) T. Higuchi, K. P. Patel, E. R. Bonow, and J. Ladsman, J . Am.

(5) I. J. Holcomb, R. B. Luers, and S. A. Fusari, J . Pharm. Sci., 62,150.4

(6) P. J. Twitchett and A. C. Moffat, J . Chromatogr., 111, 149

(7) S. E. Hays, L. T. Grady, and A. V. Kruegel, J . Pharnz. Sci., 62,1509

Co., Easton, Pa., 1975, p. 332.

London, England, 1973, p. 331.

Pharm. Assoc., Sci. Ed., 41, 293 (1952).

(1973).

(1975).

(1973).

Journal of Pharmaceutical Sciences I 151 1 Vol. 67, No. 11. November 1978